Testing the monthly anomaly with stochastic dominance

Abstract

Purpose

This study uses stochastic dominance (SD) theory to examine whether the traditional festival, such as the Spring Festival (often in February), affects the patterns of monthly anomaly for the Taiwan Stock Exchange (TWSE). The paper aims to discuss these issues.

Design/methodology/approach

The authors employ a new bootstrap-based test due to Linton, Maasoumi and Whang (hereafter LMW). The LMW test is well suited for financial time series data, such as monthly returns of various portfolios in this study, because it allows for general dependence among the prospects (distributions) and does not require the observations to be identically and independently distributed.

Findings

The particular findings of this study are that the February effect and the February-size effect indeed exist in the TWSE. Furthermore, allowing part of investors' assets is invested in the risky asset and the remaining part in a risk-free asset, first finding for monthly anomaly in the extant literature, is useful in distinguishing the performance among various size-month portfolios.

Originality/value

Instead of tax-loss and window dressing hypothesis, the Spring Festival money movement hypothesis can be used to well explain the findings.

Citation

Ke, M.-C., Chou, J.-H., Hsieh, C.-S., Chi, T.-L., Chen, C.-T. and Liang Liao, T. (2014), "Testing the monthly anomaly with stochastic dominance", Managerial Finance, Vol. 40 No. 2, pp. 137-156. https://doi.org/10.1108/MF-07-2013-0182